| Research Article |
Open Access |
|
| Tor Brunzell1*, Sören Holm2 and Bengt Jonsson2 |
| 1School of Business, Stockholm University, Sweden |
| 2Swedish University of Agricultural Sciences, Umeå, Sweden |
| *Corresponding authors: |
Tor Brunzell
Stockholm University
School of Business, SE-106 91 Stockholm
Sweden
E-mail: tb@fek.su.se |
|
| Â |
| Received Spetember 25, 2012; Published November 30, 2012 |
| Â |
| Citation: Brunzell T, Holm S, Jonsson B (2012) A Model of Optimal Dividend Policy to Maximize Shareholder Wealth When Taxes are Considered. 1:503. doi:10.4172/ scientificreports.503 |
| Â |
| Copyright: © 2012 Vuai SAH. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. |
| Â |
| Abstract |
| Â |
| This article theoretically analyzes how a firm maximizes the value of shareholders’ wealth with its dividend policy. Corporate dividend policy is one of the major puzzles in modern finance. The overall question is whether a company should pay out a dividend at all. However, the large majority of listed companies pays a dividend and also carries sophisticated dividend policies. In this paper, we outline when it is optimal for a company to pay out a dividend and when it should reinvest the profit from operations. The model takes taxes into consideration by estimating the value of a company, i.e., its present value after tax deductions, which is used as an objective function. Four different taxes are considered. The analysis shows the terms on which it is profitable to receive a dividend payout or to reinvest at an arbitrary time. Under the assumption of a unique maximum net present value, the terms at the time for the maximum net present value are also presented. |
| Â |
| Keywords |
| Â |
| Dividend; Financing; Management; Reinvestment; Taxes |
| Â |
| JEL Code |
| Â |
| C61, E62, G31, G35 |
| Â |
| Introduction |
| Â |
| One of the main questions of modern finance is how a company’s dividend policy affects shareholders’ wealth [1]. In this study, we construct a model that assumes the main assumptions of Miller- Modigliani’s [2] classical articles of dividend policy, with the addition of taxes. The aim is to analyze how a firm maximizes the value of shareholders’ wealth with its dividend policy versus the reinvestment of the profits from operations when taxes are considered. This is equivalent to maximizing the market value of the firm. Consequently, the netgp present value, i.e., the present value after any deductions for taxes, is used as an objective function in the study. |
| Â |
| The basic idea of a dividend was initially to make equity seem like a loan, where the dividend would give the equity investor a tangible return, thereby enabling a value calculation of shares [3]. Dividends would then be comparable to each other, and stock dividends could consequently be viewed as an indicator of company value, in that it shows the relationship between cash-flow and price. The dividend issue was treated early on by many economic scientists including Lintner [4], Gordon [5-7], Miller and Modigliani [2], Farrar and Selwyn [8-12]. |
| Â |
| Since the seminal theoretical paper by [2], in which the authors showed the irrelevance of a dividend in a frictionless world, the obvious basis regarding a dividend is why companies should pay it out at all. However, the vast majority of companies pay dividends, and they also apply a sophisticated dividend policy. Why the real world is not like the academic model is not explained by the Miller-Modigliani (MM) irrelevance argument. However, there are several explanations complementing the MM irrelevance argument given by the academic literature: the signaling theory [13,14] claims that a dividend is a means to communicate company information to its shareholders about the company’s future; the agency theory [15,16] claims that dividends are rooted in the conflicting interests within a company of different stakeholder groups - stockholders, management, and bondholders; the clientele theory [17,18] claims that a dividend is paid out to satisfy the payout demands from a heterogeneous dividend clientele; and the catering theory [19,20] claims that stock market mispricing might influence individual firms' investment decisions. Here, managers cater to investors by paying dividends when investors prefer dividend payers and omitting dividends when investors prefer non-payers. |
| Â |
| Surveys of dividend policy provide a good understanding of the minds of the people responsible for companies’ dividend policies. In 1956, a survey study by Professor John Lintner detailed what is considered to be the foundation of modern understanding of dividend policy. The study was conducted in detail by interviewing the managers directing the dividend policies of 28 companies. The main findings were that managers used long-run target payout ratios to determine dividends, and they changed the company's dividends only when they believed that the company’s permanent earnings had changed. Since then, several researchers have used the survey method to learn more about the factors that influence the construction of dividend policy [21,22]. In a study by Brav [23] the authors compared the treatments of dividends and stock repurchases. The study was performed by surveying financial executives with an instrument that focuses on both dividends and repurchases. The study was further developed by conducting one-on-one interviews with a number of CFOs. The findings showed that dividend policy is still very conservative, while stock repurchase is more flexible. Additionally, it was shown that companies perceived large penalties for cuts and only small rewards for increases. |
| Â |
| The structure of this article is as follows: in Section 2, we present the most important and relevant research on dividend policy; in Section 3, we present the model; in Section 4, we give an analysis of the model; in Section 5, we present an example of a practical case based on the model; and in Section 6, we draw the conclusions of the study. |
| Â |
| Previous Research |
| Â |
| The tax burden on dividends is due partially to the tax on company profits and the tax treatment of the dividend recipient. To understand the complexity of dividend policy, the taxes on dividends should be compared to the effective capital gains tax. Typically, when the tax on a dividend exceeds the tax on capital gains, companies will prefer to retain earnings or to repurchase shares [24-26]. |
| Â |
| In principle, research shows that taxes affect dividend policy. However, exactly how much impact taxes have on dividend policy is a matter of debate among scholars. Brav [23] and [27] argue that taxes do not have a dominant effect on the dividend decisions for most companies. In contrast, [28] argue that taxes can have a significant impact on corporate dividend decisions. Alzahrani and Lasfer showed that companies in countries with a double taxation tax system pay significantly lower dividends than companies in countries with partial-imputation systems. Because taxes are different depending on the type of shareholders, corporate shareholders and institutional investors prefer dividends before capital gains, whereas smaller shareholders prefer capital gains. Therefore, the large shareholders may use their power to directly influence dividend payouts for tax reasons [17]. |
| Â |
| Historically, capital gains have been taxed at a lower rate than dividends, which means that investors often prefer capital gains over dividends. Furthermore, capital gains are not taxed until shares are actually sold; therefore, investors can control when capital gains are realized and when taxes are paid. Investors, however, cannot control dividends because it is the board of the related company that decides on the distribution. Many scholars have evaluated the tax effect on dividend policy from the tax change that the U.S. Congress passed in 2003. According to the adopted legislation, dividends are taxed at the same rate as long-term capital gains. Studies, such as those of Chetty and Saez [29-31], have found significant increases in dividend payouts and initiations following the adoption of the legislation. Chetty [29] found that the aggregate regular dividends increased by 20 % within 1.5 years of the introduction of the legislation, prompting the authors to conclude that the tax cuts caused this increase. |
| Â |
| In corporate finance, a finance manager generally has two operational decision areas to consider: the investment (or capital budgeting) and the financing decisions. The capital budgeting decisions concern the choice of what long-term assets the company should acquire, while the financing decisions concern how these assets should be financed. A third decision area arises, however, when the company begins to generate profits. How much of the earnings should the company distribute as dividends to shareholders, and how much should the company reinvest in the business? The ultimate goal of managers should be the maximization of shareholder wealth, so every action taken by the managers should consequently be to maximize the wealth of the shareholders. The managers do not just have to consider how much of the company’s earnings are needed for investments, but they also have to consider the effect of their decisions on shareholder wealth [32]. |
| Â |
| Shareholder wealth is defined as the present value of expected future returns to shareholders. These returns are either regular dividend and/or represent revenue from the sale of shares. Shareholder wealth is measured as the market value of the company's shares. The goal of maximizing shareholder wealth is a long-term goal, where shareholder wealth is a function of all future monetary returns to the shareholders. To make decisions that maximize shareholder wealth, management must take into account the long-term effects on the company and not just focus on short-term effects [33]. |
| Â |
| Model |
| Â |
| The starting point in building our tax-including model is the assumptions of [2]; the authors assumed an ideal economy characterized by rational behavior and perfect certainty (deterministic approach) and to some extent under risk (stochastic perspective). Their proposition also assumed that no transaction costs exist and that individuals and corporations borrow at the same rates. A limitation in the analysis is that the sum of funds used for dividends and investments is equal to the profits from operations. Thus, neither external financing nor share repurchase is considered. |
| Â |
| In their 1986 article, Shleifer and Vishny discuss the dividend puzzle using four taxes. In this article we consider four taxes similar to the ones outlined by Sheifer and Vishny: a) corporate tax when dividend payout is received; (b) shareholder´s personal income tax on dividends; (c) corporate tax when reinvestment is made; and (d) shareholder´s capital gains tax. |
| Â |
| Concepts and notations |
| Â |
In this study, the net present value, i.e., the present value after deduction for taxes, is used as an objective function. In accordance with this, total net present value over time t, PV(t), is the sum of the net present values of dividends PVD(t) and gains of reinvestments PVR(t).
Thus, PV(t)= PVD(t)+ PVR(t). |
| Â |
| In the study, the following concepts and denotations are used. |
| Â |
| ´Gross´ means ´with no deduction for taxes´; ´net´ means ´after deduction for taxes´. |
| Â |
| t denotes an arbitrary terminal time point, |
| Â |
| s denotes time point ending period s in the discrete model; time in the continuous model where 0 ≤ s ≤ t, |
| Â |
| G(s) denotes the shareholder´s gross profit from operations in period s; it is assumed to be non-negative, |
| Â |
| q(s) denotes dividend ratio, i.e. relative dividend fraction of the firm’s gross profit at time point s where 0 ≤ q(s) ≤ 1; q is decision function, |
| Â |
| 1-q(s) denotes retention ratio, i.e. relative reinvestment fraction of the firm´s gross profit at time point s where 0 ≤ q(s) ≤ 1, |
| Â |
| r(s) denotes market-required rate of return in period s (interest rate); same notation in both the discrete and continuous models, but for consistency with slightly different values, |
| Â |
| δ(s) denotes discount factor in period s in the discrete model, i.e., |
| Â |
 |
| Â |
| g(s) denotes rate of return on reinvestment in period s (growth rate); same notation in the discrete and continuous model, but for consistency with slightly different values |
| Â |
| γ(s) denotes growth factor in period s, the discrete model, γ(s)=1+g(s). Different tax rates, c, all 0≤τi<s. |
| Â |
| Ï„1(s) denotes corporate tax rate on gross profit in period s, when dividends are received, |
| Â |
| Ï„2(s) denotes personal income tax rate on gross dividend payout in period s, |
| Â |
| Ï„3(s) denotes corporate tax rate on gross profit in period s, when reinvestment is made, |
| Â |
| τ4(s) denotes capital gains tax rate, i.e. personal income tax rate in period s on the capitalized reinvestments after a deduction for corporate tax; τ4(s)τ4(t)= for all S ≤ t where t is sale date |
| Â |
| PV(t) denotes total net present value over t periods, |
| Â |
| PVD (t) denotes net present value of dividends over t periods, |
| Â |
| PVR (t) denotes net present value of capital gains of reinvestments over t periods. |
| Â |
| Net present value of dividends |
| Â |
| The net dividend in period s equals G(s).q(s).(1-Ï„1(s)) (1-Ï„2(s)) for s = 0,1... |
| Â |
| By discounting and summing over s we obtain |
| Â |
 |
| Â |
| where δ(0)=1. |
| Â |
| For convenience, we now make the discrete expression (1) continuous in time; thus, a change occurs from discontinuous to continuous time. Then, δ(u) is replaced by the continuous version exp(-r(u)). We obtain |
| Â |
 Â Â Â Â Â Â Â Â Â Â Â Â (2) |
| Â |
| Net present value of reinvestments |
| Â |
| The net reinvestment gain at time point s equals G(s).(1-q(s)).(1-τ3(s)). This is increased by future growth up to time t, where it is taxed by the ratio τ4(t) and discounted back to time s=0. With the convention τ4(s)= τ4(t) for 0 ≤ s ≤ t we obtain |
| Â |
 |
| Â |
| As above, a change is made from discrete to continuous time, and simultaneously δ(u) and γ(u) are replaced by their continuous analogues r and g. We obtain |
| Â |
 |
| Â |
| Net present value of both dividends and reinvestments |
| Â |
| Adding formulas (2) and (4) results in |
| Â |
 |
| Â |
| Analysis |
| Â |
| Rearranging (5) by assembling all terms that contain the decision function q we obtain |
| Â |
 |
| Â |
| Consider PV(t) as a functional of q(.); all other functions are given. |
| Â |
| are non-negative, it is found that PV(t) is maximized by choosing q(s)=t whenever h(s)>0, and q(s)=0 whenever h(s)i:s fulfill 0 ≤ τi <1, the optimum for an arbitrary time s is obtained by choosing |
| Â |
 |
| Â |
 |
| Â |
| total rate of return on reinvestment during the period s to t, and rT(s,t) is the corresponding total market-required rate of return (taxes not included). The time length t can be chosen arbitrarily but fixed. |
| Â |
| Taking the derivative of PV(t) with respect to t (in points of differentiability), we obtain |
| Â |
 |
| Â |
| The first term is the present value of the sum of net dividends and net reinvestment at t. PVR(t) in the second term is the present value of the accumulated, capitalized net reinvestments up to t (see (4)), g(t).PVR(t) is the present value of the revenue of this reinvested capital, and r(t).PVR(t) is the present value of the cost to keep this capital. At optimum, q(t) is either 0 or 1, and the first term is then the present value of either the net dividend (q(t)=1) or the net investment (q(t)=0 , in both cases after taxes. We now extend our model to include the assumption that the rate of return on reinvestment in period s depends on the size of the reinvestment. To do this, G(s) is divided into monetary units ordered from 0 and upwards. The rate of return at time t of the vth unit reinvested at time s is denoted g(s,t,v), and the dividend ratio of the vth unit at time s is denoted g(s,v). |
| Â |
| Then, (6) is extended as follows |
| Â |
 |
| Â |
 |
| Â |
| In the same way as above, it is observed that PV(t) is maximized by choosing |
| Â |
 |
| Â |
| If g and r are continuous functions and g is non-increasing, the solution (9) gives an upper limit ν*= ν*(s,t)for optimal reinvestment. |
| Â |
| So far all functions are considered to be known. Let this assumption be relaxed by assuming that the growth and interest rates are random variables and that we want to maximize the expected value of the present value PV(t). The same arguments as above lead to the same decision rules (7) and (9), but with the left hand side replaced by its expected value. Because by Jensen’s inequality [34] the expected value E(exp{U}) ≥ exp{E(U)} for any random variable U, q(s)=0 more often than expressed by inserting the expected rates in the rules (7) and (9). The expectation is then taken conditionally on the r and g processes up to time s. Thus, the random rates of growth and interest should increase the retention rate compared to known rates if the expected rates are the same. |
| Â |
| An Example of a Practical Case |
| Â |
| Above, we have theoretically shown the “equilibrium†terms on which it is profitable to change from receiving dividends to reinvestment or vice versa, when four different taxes are considered. In the simple example below, we will calculate the average annual rate of return on reinvestment - under the above-mentioned terms - during a ten-year period under the following hypothetical terms: |
| Â |
| - corporate tax rate on gross profit (Ï„1) is 28 % when a dividend is received, |
| Â |
| - personal income tax rate on gross dividend payout (Ï„2) is 30 %, |
| Â |
| - corporate tax rate on gross profit (Ï„3) is 20 % when reinvestment is made, |
| Â |
| - capital gains tax rate (Ï„4) is 20 %, |
| Â |
| - annual market-required rate of return is 5 % during the period, and |
| Â |
| - annual rate of return on reinvestment is g % during the period. |
| Â |
| According to the discrete version of the solution (7), we obtain the breakpoint between q=0 and q=1 when |
| Â |
 |
| Â |
| which gives g=2.521... |
| Â |
| In this case, the average annual rate of return on reinvestment is 2.5 % during the period and is thus equivalent to a market-required annual rate of 5 %. |
| Â |
| Conclusions |
| Â |
| Assuming that the sum of funds used for dividends and reinvestments is equal to the profits from operations, we find from (7) that the value of the shareholders’ wealth (i.e., equivalent to the value of the firm) is affected by dividend policy in a world both with and without taxes. Dividend/reinvestment policy is irrelevant in this respect |
| Â |
 |
| Â |
| tax rates; gT(s,t) is the total rate of return on reinvestment during the period s to t; and rTT(s,t) is the corresponding total market-required rate of return. |
| Â |
| We find from (7) that it is advantageous for shareholders to receive dividend payouts at time s, when the ratio is less than the ratio |
| Â |
 |
| Â |
| Otherwise, if this relationship is larger, it is better to reinvest. |
| Â |
| However, in the case of random rates of growth and interest, the retention rate should be increased compared to the known rates of return, if the expected rates are the same. |
| Â |
| The derivative of PV(t) (see (8)) set equal to zero determines the relationship between the included components at the time of maximum net present value. Here we assume that the present value, as a function of time, has a unique maximum. Then, the time for the maximum net present value occurs when the sum of the net dividends and the net reinvestment is equal to the difference between the cost to keep the accumulated net reinvested capital and the revenue of the same capital. |
| Â |
| |
| References |
| Â |
- Black F (1976) The dividend puzzle. J Portfolio Manage 2: 5-8.
- Miller MH, Modigliani F (1961) Dividend Policy, Growth and the Valuation of Shares. J Bus 34: 411-433.
- Frankfurter GM, Wood BG (1997) The evolution of corporate dividend policy. Journal of Financial Education23: 16-32.
- Lintner J (1956) Distribution of Incomes of Corporations among Dividends, Retained Earnings and Taxes. Am Econ Rev 46: 97-113.
- Gordon MJ (1959) Dividends, Earnings, and Stock Prices. Rev Econ Stat 41: 99-105.
- Gordon MJ (1962). The Savings Investment and Valuation of a Corporation. Rev Econ Stat 44: 37-51.
- Gordon MJ (1994) Finance, investment and macroeconomics: the neoclassical and post Keynesian solution. Aldershot, U.K.: Elgar.
- Farrar DE, Selwyn LL (1967) Taxes, Corporate Financial Policy and Return to Investors. National Tax Journal 20: 444-454.
- Fama EF (1974) The Empirical Relationships between the Dividend and Investment Decisions of the Firms. Am Econ Rev63: 304-318.
- Brennan MJ (1970) Taxes, Market Valuation and Corporate Financial Policy. Natl Tax J 23: 417-427.
- Miller MH, Scholes M (1978) Dividends and taxes. J Financ Econ, Vol. 6: 333-364.
- Elston JA (1996) Dividend policy and investment: Theory and evidence from US panel data. Managerial and Decision Economics 17: 267-275.
- Bhattacharya S (1979) Imperfect Information, Dividend Policy, and “The Bird in the Hand” Fallacy. Bell Journal of Economics10: 259-270.
- Miller MH, Rock K (1985) Dividend Policy under Asymmetric Information. J Financ 40: 1031-1051.
- Easterbrook FH (1984) Two Agency-Cost Explanations of Dividends. Am Econ Rev74: 650-659.
- Jensen MC (1986) Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers. Am Econ Rev76: 323-329.
- Shleifer A, Vishny R (1986) Large Shareholders and Corporate Control. J Polit Econ 94: 461-488.
- Allen F, Bernardo AE, Welch I (2000) A Theory of Dividends Based on Tax Clienteles. J Financ 55: 2499-2536.
- Baker M, Wurgler J (2004a) A Catering Theory of Dividends. J Financ 59: 1125-1165.
- Baker M, Wurgler J (2004b) Appearing and Disappearing Dividends: The Link to Catering Incentives. J Financ Econ 73: 271-288.
- de Jong A, van Dijk R, Veld C (2003) The dividend and share repurchase policies of Canadian firms: empirical evidence based on an alternative research design. International Review of Financial Analysis 12: 349-377.
- Baker HK, Powell GE, Veit ET (2002) Revisiting the dividend puzzle: Do all of the pieces now fit? Review of Financial Economics 11: 241-261.
- Brav A, Graham JR, Harvey CR, Michaely R (2005) Payout Policy in the 21st Century. J Financ Econ 77: 483-527.
- Auerbach AJ (1979) Wealth Maximization and the Cost of Capital. Q J Econ 93: 433-446.
- Bradford DF (1981) The Incidence and Allocation Effects of a Tax on Corporate Distributions. J Public Econ 15: 1-22.
- Auerbach AJ, Hassett KA (2007) The 2003 Dividend Tax Cuts and the Value of the Firm: An Event Study. In A. Auerbach, J. Hines, and J. Slemrod, eds., Taxing Corporate Income in the 21st Century. Cambridge: Cambridge University Press 93-126.
- Julio B, Ikenberry DL (2004) Reappearing Dividends. Journal of Applied Corporate Finance 16: 89-100.
- Alzahrani M, Lasfer M (2012) Investor protection, taxation, and dividends. Journal of Corporate Finance 18: 745-762.
- Chetty R, Saez E (2005) Dividend Taxes and Corporate Behavior: Evidence from the 2003 Dividend Tax Cut. Q J Econ 120: 791-833.
- Brown JR, Liang N, Weisbenner S (2007) Executive Financial Incentives and Payout Policy: Firm Responses to the 2003 Dividend Tax Cut. J Financ 62: 1935-1965.
- Blouin JL, Raedy JS, Shackelford DA (2011) Dividends, Share Repurchases, and Tax Clienteles: Evidence from the 2003 Reductions in Shareholder Taxes. Account Rev 86: 887-914.
- Bishop SR, Crapp HR, Faff RW, Twite GJ (2000) Corporate Finance. Prentice Hall Inc., Sydney.
- Damodaran A (2006) Damodaran on Valuation. John Wiley & Sons Inc. New York, N.Y.
- Feller W (1966) An Introduction to Probability Theory and Its Applications. Volume II,
John Wiley & Sons, New York.
|
| Â |
| Â |
